Issue/Introduction

This article covers a ZStone 4.2.8 host addition failure in which the node operating system changes from CentOS 7.6 to Helix 7.6 after install.sh runs successfully and the host is added through the UI.

Symptoms

• install.sh runs successfully.
• Adding the ZStone host through the UI fails.
• A background inspection shows that the node operating system has changed from CentOS 7.6 to Helix 7.6.

Environment

• Product line: ZCF
• Component: ZStone
• Version: 4.2.8
• Operating system: CentOS 7.6

Possible causes

When a ZStone host is added, the ZStack Cloud Ansible service installs the Helix RPM, which replaces the node operating system with Helix and causes the host addition to fail.

Resolution

1. Inspect the apiserver log

• Check: Confirm whether yum installation records for the ZStone product exist during the host addition.
• If abnormal: Continue to inspect the RPM dependencies and yum installation order.

2. Verify the RPM dependencies

• Check: Try to install the product RPMs manually and confirm whether dependencies handled automatically by yum trigger the Helix package installation.
• If abnormal: Record the affected packages and their dependencies if installing the product RPMs also triggers the Helix package installation.

3. Inspect the yum and Helix records

• Commands:

  cat /var/log/yum.log
  grep -Rn "helix" /var/log/*

• Check: Confirm the installation order, timestamps, and related operations for the Helix packages.

4. Confirm the source of the issue

• Check: Confirm that the Helix package installation is related to the ZStack Cloud Ansible service.
• If abnormal: Escalate the issue to the ZStack Cloud team so that the product-level host addition process can be corrected to avoid triggering the Helix package installation.